The proposed study is a novel analysis of a large clinical trial which brings innovative sequencing approaches and novel investigation of gag gene mutation to determine the mechanism of drug resistance and failure among recipients of a boosted protease regimen, Atazanavir and ritonavir (ATV/r). We seek to understand the mechanism(s) of virologic failure among Atazanavir/ritonavir (ATV/r) recipients where we have clear evidence that adherence and pharmacokinetics are insufficient to explain virologic failure without consensus protease inhibitor mutations. The hypotheses to be explored include the investigation of minority quasispecies through cloning and ultra deep sequencing of protease to identify low levels of Protease inhibitor drug resistance mutations in the the gag gene. Alternatively, the protease and gag and gag cleavage sites (where protease cleaves the polyprotein products of HIV replication) may mutate to compete effectively with protease inhibitors. We are proposing a pilot study to determine mechanism(s) which may explain virologic failure following a boosted PI regimen. By determining the mechanism of virologic failure it will be possible to design diagnostic sequencing to identify the optimal sequential regimen(s) likely to be effective. The virologic explanations, which would have an important impact on subsequent diagnostic and treatment strategies include 1) the presence of multiple minor and major PI drug resistance mutations at levels below those detected in consensus sequencing (< 20%). Alternatively, there may be mutations selected at protease cleavage sites as mutations in the gag-pol polypeptides which allow the wild-type protease substrates to effectively compete with the inhibitor (Atazanavir) with continued replication of wild-type virus. There are implications for future treatment in either case. If there are multiple, minority PI resistance or cleavage site mutations, uncovered, then these patients would require different drugs, or drug class to achieve virologic suppression. Additionally, we will also determine whether there are mutations in the reverse transcriptase (such as M184V, K65R, or L74V) at levels that cannot be detected by consensus genotyping (<20%) but which as minority species, jeopardize a new regimen including those with a low genetic barrier to resistance such as an NNRTI (Efavirenz) or an integrase inhibitor (Raltegravir). We hypothesize that the mechanism of failure and the selection of mutations among boosted protease inhibitor recipients will point to the most effective regimen(s) for subsequent treatment. We will bring new, sensitive and innovative methodologies to bear on these questions; Ultradeep pyrosequencing and multiple cloning methods can detect low levels of mutant viruses (< 1%). This study will provide robust data towards the development of a randomized trial of new diagnostic approaches and regimens for ATV/r failures, after one or more screening tests for minority quasi-species and/or cleavage site mutations, each of which may reduce the activity of a new drug regimen. The overall aim is to enhance drug treatment strategies by examining virologic failure samples from a large randomized trial to test each of the hypothesized mechanisms for virologic failure. Data on the selection of drug resistance through minority quasi-species PI resistance mutations, gag-pol and env cleavage site mutations or both occurring below the levels of detection of standard genotyping will be key to the development of a randomized diagnostic and treatment trial to determine the best approach to ATV/r failures.